Composition and Seasonal Phenology of a Nonindigenous Root-Feeding (Coleoptera: ) Complex in Northern Hardwood Forests in the Great Lakes Region

Environ. Entomol. 34(2) : 298-307 (2005) ABSTRACT oblongus (L.), sericeus (Schaller), and asperatus (Bonsdorff)comprise a complex of nonindigenous root-feeding in northern hardwood forests of the Great Lakes region. Little is known about their detailed biology, seasonality,relative abundance, and distribution patterns. We studied 10 sites over a 2-yr period. Two sites were in northeastern Wisconsin, and eight were in the neighboring southern upper peninsula of Michigan. Larval abun- dance was estimated by soil sampling, and adult abundance was estimated by sweep netting, emer- gence trapping, and beating samples. Sweep netting collected the most weevils overall (71.0%), whereas beating and emergence traps collected 22.1 and 6.9%, respectively. P. sericeus were the predominant larvae, representing 34.3% of total Curculionidae, whereas P. oblongus were the pre- dominant adults, representing 66.4% of Curculionidae. Few S. asperatus and Trachyphloeus aristatus (Gyllenhal) larvae and adults were collected, with the latter being a new record for Wisconsin. Two additional species, pellucidus (Boheman) and an undetermined Polydrusus sp., were collected only as larvae. Six species of curculionids were collected overall, with at least five being confirmed as nonindigenous species. P. oblongus and P. sericeus adults were the most abundant. These did not coincide temporally. Over 63% of P. obhgus and P. sericeus were collected during single 4-wk intervals in mid-June and mid-July, respectively. Conversely, S. asperatus adults overlapped with both other species, occurring sparingly from 4 June through 28 August. One species was predominant at each site and generally accounted for 180%of the total weevil population. P. oblongus larvae and adults predominated in five and eight sites, respectively, whereas P. sericeus and S. asperatus larvae and adults predominated in one site each. Adult and larval populations were generally clustered. We evaluated vertical stratification of P. sericeus larvae in the soil, and most were located within the top 10 cm.

KEY WORDS Phyllobius, Polydrusus, Sciaphilus, root herbivores, invasive species

INVASIVE NONINDIGENOUS FOREST INSE~are estimated to losses is particularly poorly characterized (Wilcove et cost the United States $2.1 billion annually in losses to al. 1998, Lodge and Shrader-Frechette 2003). forest products, excluding eradication and control ex- Past research has focused more heavily on invasive penses (Pimentel et al. 2000). As of 1994, -360 non- in production than natural ecosystems, such as indigenous species were established in North deciduous forests. Of the latter, we know relatively American forests, and 30% of these were serious more about the biology and impact of folivores such as economic or ecological threats (Mattson et al. 1994, the gypsy moth (Lepidoptera: Lymantriidae: Lyman- Pimentel et al. 2000). Potential ecological disturbance tria dispar L.) and larch casebearer [Lepidoptera: by invasive species includes loss of native , Coleophoridae: Coleophora laricella (Hiibner) ] species extinction, interspecific competition, preda- (Liebhold et al. 1995), subcortical insects such as the tion, and alteration of ecosystem processes (Vitousek pine shoot (Coleoptera: Scolytinae: Tomicus et al. 1996, Wilcove et al. 1998, Mack et al. 2000, Sala piniperda L.) (Haack and Poland 2001) and Asian et al. 2000). The extent of ecosystem disturbance by longhorned beetle [Coleoptera: Cerambycidae: Ano- nonindigenous species that do not cause commercial plophora glabripennis (Motschulsky) ] (Nowak et al. 2001). and sap feeding; insects such as the hemlock woolli adelgid (~emiitera:Adelgidae: Adelges tsugae Department of Entomology, University of Wisconsin-Madison, Armand) (McClure and Cheah l999) and balsam 345 Russell Laboratories, 1630 Linden Dr., Madison, WI 53706. woolly adelgid [Hemiptera: Adelgidae: Adelges piceae Corresponding author: Wisconsin Department of Natural Re- (Ratzeburg)] (Hain et al. 1991), than root-fee&ng sources, 3911 Fish Hatchery Rd., Fitchburg, WI 53711 (e-mail: [email protected]) . (H~~~~~2001). hi^ is largely because of dif- 3 USDA Forest Service, Forestry Sciences Laboratory, 5985 High- ficultie~in studying root herbivores in general and way K, Rhinelander, M~I54501. because below-ground herbivores exert subtle effects

0046-225Xl0510298-0307$04.00/0 O 2005 Entomological Society of America April 2005 PINSKIET AL.:ABUNDANCE AND SEASONA~OF NONINDIGENOUS WEEVILS 299 that may go unnoticed for substantial periods (Hunter plant records of P. sericeus in North America include 2001). Potential negative impacts of below-ground maple, poplar (Populus spp.), and hazel (Corylus herbivores include reduced plant diversity, altered americana Walter) (Frost 1946). In Europe, they successional processes, altered carbon and nitrogen have been observed feeding on alder (Alnus spp.), allocation between roots and foliage, increased sus- hazel (Corylus avellana L.), willow (Sabx americana ceptibility to other herbivores and pathogens, and Walb.) ,and a variety of fruit trees and shrubs (Parrott reduced total yield (Syvertsen and McCoy 1985, and Glasgow 1916, Morris 1978, Gharadjedaghi 1997, Brown and Gange 1991,1992, Hunter 2001). Czerniakowski 1998). S. asperatus has been observed Three root-feeding weevils (Coleoptera: Curcu- feeding primarily on maple, birch, and various fruit lionidae) of European origin are currently established trees and shrubs (Henshaw 1888, Witter and Fields 1977, in northern deciduous forests of the Great Lakes region: Levesque and Levesque 1994). In general, P. obhgus (L.), Polydrusus sericeus (Schaller), and S. asperatus seem to feed more heavily on sugar and (Bonsdorff).P. oblongLls were maple (Acer sacchamm Marshall),whereas P. sericeus has &st collected in New York in 1923 from American elm been reported feeding in northern hardwood stands in (Ulmus ammicana L)(Felt 1928);S. asperatus were first which maple is not the dominant tree species (Witter collected in Sydney, Nova Scotia, in 1884 (Brown 1940); and Fields 1977). and P. sericeus were first collected near Indianapolis, IN, There are few reports of damage by these weevils before 1916 (Blatchley and Leng 1916).All three species under natural conditions in Europe, but they can be occur throughout the northeastern United States and important in commercial settings. Adults of P. oblongus adjacent Canada and as far west as Minnesota (O'Brien are pests of pear (Pyw spp.) and apple (Malus spp.) and Wibmer 1982, Downie and Arnett 1994, Anderson orchards throughout Great Britain and the Nether- 2002). S. asperatus has also been found in South Dakota, lands (Massee 1932, Helsen and Blommers 1988); lar- Idaho, Quebec, and British Columbia (Witterand Fields val P. sericeus are considered a serious root-feeding 1977, Levesque and Levesque 1994, Anderson 2002). pest in young nursery platings of oak (Qwmspp.) The life histories of these weevils have been re- and hazel (Corylus spp.) in Belgium (Casteels and De ported in general terms from their native European Clercq 1988); larval and adult S. asperatus are impor- ranges. They have similar life cycles, with the excep- tant pests attacking strawberry (Fragariaspp.) roots in tion that P. oblongus and P. sericeus reproduce sexually, Great Britain (Massee 1954, Ministry of Agriculture, whereas S. asperatus is parthenogenic and triploid Fisheries, and Food 1963) and raspberry (Rubus spp.) (Suomalainen et al. 1987).P. oblongus and P. sericeus leaves in Russia (Witter and Fields 1977). Past re- are diurnal feeders that disperse by flight, whereas search in the Great Lakes region has led to speculation S. asperatus is a nocturnal feeder and flightless. Adults that twig scarring and subsequent forked branching of emerge from the soil in spring and seem to be rela- sugar maple resulted from P. oblongus and S. asperatus tively polyphagous on foliage of deciduous trees and feeding (Simmons and Knight 1973, Witter and Fields understory vegetation (Parrott and Glasgow 1916, 1977). Hoffman 1950, Massee 1954, Vollman 1954, Fields Little research on this nonindigenous weevil com- 1974). It has been speculated that they may also feed plex has been conducted in North America. Thus we on flower buds and blossoms, leaf buds, newly sprout- have incomplete knowledge of their basic biology, sea- ing leaves, and terminal shoots (Massee 1954, Fields sonality, relative abundance, and distribution patterns. 1974).A notch-like wound around the leaf perimeter The objectives of this study were to (1) determine rel- characterizes adult feeding. Copulation begins imme- ative abundance of larval and adult P. oblong.us,P. sericeus diately on emergence and occurs at the feeding site. and S. asperatus populations, (2) determine the seasonal Oviposition occurs 10-14 d later, 2-5 cm beneath the phenologies of larvae and adults, and (3) determine the soil surface (Vollman 1954). Larvae emerge -30 d vertical distribution of P. sdlarvae in the soil. later and burrow into the soil in search of fine root material. P. oblongus has five larval instars (Vollman Materials and Methods 1954),but the number for P. sericeus and S. asperatus is undetermined. Larvae overwinter in the soil as late Site Selection. Ten sites were selected in mesic instars before pupation occurs the following spring. northern hardwood stands throughout northeastern There are also reports of S. asperatus completing de- Wisconsin and the Upper Peninsula of Michigan, dur- velopment by fall and overwintering as adults in leaf ing autumn 2001 (Table 1).Sites were selected based litter, thus resulting in two spring emergence peaks on previous observations of weevil populations and (Ministry of Agriculture, Fisheries, and Food 1963, availability of detailed data on vegetative composition Levesque and Levesque 1994). and soil parameters from previous ecosystem studies Observed host plants of P. oblongus in North Amer- (Goodburn 1996, Bockheim 1997, Goodburn and ica include a wide variety of tree and shrub species, Lorimer 1998,1999).Climate in this region is charac- particularly maple (Acer spp. ) , birch (Betula spp.) , terized by an average temperature range of -12 to hophornbeam [Ostrya virginiana (Miller) K. Koch], lg°C, annual frost-free period of 100 d, and 850 mm of aspen (Populus spp.), willow (Salix spp.), American precipitation (Goodburn 1996). Soil texture is classi- basswood (Tilia americana L.), American elm ( Ulmus fied as sandy loam in 8 of the 10 sites and as silt in 2 americana L.), and several fruit trees, berries, and sites (Bockheim 1997).The sites included a mixture of shrubs (Felt 1928, Carruth 1936, Fields 1974). Host old growth forests, recently thinned forests, and a ONMENTAL ENTOMOLOGY Vol. 34, no. 2

hybrid poplar plantation (Table 1).Sugar maple is the predominant tree species in 8 of the 10 sites; red maple (Acer mcbrum L.) and hybrid poplar (Populus spp.) are each predominant in 1 site (Table 1).Other common trees include yellow birch (Betula alleghaniensis Brit- ton), paper birch (Betula papyrifera Marshall),Amer- ican basswood (Tilia americana L.), northern red oak ( Quercus rubra L. ) ,white ash (Fraxinusamericana L. ) , black Ash (Fraxinus nigra Marshall), ironwood [Os- trya virginiana (Miller) K. Koch], black cherry (Pmw serotinu Ehrhart), and trembling aspen (Populus tremu- loides Michaux) . One 1,600-m2main plot was established at each site and was divided into four quadrants for sampling pur- poses. Each main plot was marked at the center, and its GPS coordinates were recorded. Four of the 10 study sites, hereafter termed intensive, were chosen for more frequent sampling and detailed observations (Table 1).These were the locations with the highest population density of each weevil species, based on preliminary data collected during summer 2001. Two intensive sites were chosen for P. seriwus because it was most abundant in the hybrid poplar plantation, but it was also desirable to have a more natural forest habitat. Larval Sampling. The occurrence and density of larvae were estimated by sampling soil with an auger (8.2 cm diameter by 16 cm deep, 845 cm3).Surface leaf litter was removed before taking samples. Soil samples were collected 29 September and 10 November 2001 and 3 May and 2 November 2002. In 2001, all 10 sites % Q, were sampled, eight samples at each site, two ran- a a 3 zzzE2YYYY - domly selected from each of the quadrants. Because of 00 044 b; z G G E 6882s8888, the apparent clustered distribution of larvae, the num- kkk 0Okkkk 0 2 -a -ma E Eqaqa 3 ber of study sites was decreased to allow an increased CCC0~GCCC number of soil samples per plot in 2002. In 2002, only .a .a .a 3 3 .El .P .P .P G 8 YYYea ea rn$$ZZZZZ .-C the four intensive sites were sampled, and there were ZZZ zzzzw a 0.2.: a ea ea a+ 2 40 samples per site, with 10 per quadrant. ag$H$ha$d .-8 Stratified soil samples were collected as above, with .2ggzzgggg3 .-ad g the exception of taking samples at each of three 2 depths. For each sample, the top 0-10 cm, the next 8 2= 10-20 cm, and the next 20-30 cm of soil were col-

0, lected. Stratified samples were collected 10 November .-Y 2001 and 3 May and 2 November 2002. Two intensive -3 Em sites were sampled in 2001 with six samples per site. All C four intensive sites were sampled in May 2002, with 20 .- 8 samples per site, including five randomly selected per E quadrant. Few larvae were present in all but one site, 8 so only the latter was sampled in November 2002. .-b; Each soil sample was placed in a self-sealing plastic 8 bag in the field. Samples were stored in a walk-in j. 9 2 0 cooler (5.S°C) within 4 d of collection. Samples were z 2 + generally stored for 2-3 wk, with a maximum of 8 wk g , before larval extraction. Each sample was sifted 5 , -5.-g through two sieves having 2.38- and 0.98-mm mesh, ,- E respectively. The larger mesh opening collected root P 3 .s debris, leaf litter, and large larvae, whereas the smaller --8 a passed fine soil but collected small larvae. The con- -g tents of each screen and the collection pan were ex- .zY s3 2-3 mined for 5 min. All larvae and other macro-soil PIC u a dwelling insects were collected and fixed in boiling water before preservation in 70% ethanol. The number April 2005 PINSKIET AL.:ABUNDANCE AND SEASONAL~IYOF NONINDIGENOUS WEEVILS 301 of earthworms in each sample was also recorded, and through holes cut into the sides. There were two a representative sample was retained. Voucher spec- randomly placed traps per quadrant at each site (to- imens were deposited in the University of Wisconsin- tal = 16 and 8 for 2002 and 2003, respectively). Emer- Madison, Department of Entomology,Insect Research gence traps were checked biweekly June to August Collection. 2002 and 2003. Identification to family was determined using Law- Insects were brought to the laboratory and stored in rence (1991). Curculionidae were further identified the freezer (-10°C) until identification. Generic and to using van Emden (1952). Identification of species-level idenacations of weevils were conducted P. oblongus and S. asperatus were based on van Emden using Sleeper (1957),Brown (1965),Downie and Arnett (1952), Vollman (1954), and Axelsson et al. (1973), (1994),and Monis (1997). Familial identification for all and the status of S. asperatus as the only Sciaphilus in other insects was performed by S. J. Krauth (Curator, North America (O'Brien and Wibmer 1982, Downie University of Wisconsin-Madison, Department of and Arnett 1994, Anderson 2002). Identification of Entomology, Insect Research Collection) or using Bor- P. sericeus was validated by comparison between lar- ror et al. (1989). vae collected in the field and those reared in the The variation of adult P. oblongus abundance be- laboratory from adults. tween quadrants within a site was analyzed by a $ Abundance of P. sericeus at various soil depths was test (P < 0.05), which compared an expected uniform analyzed at the hybrid poplar plantation site accord- distribution to the observed results. Abundance ing to a split plot design with subsampling (PROC within quadrants, instead of individual samples, was GLM; SAS Institute 1999). The mean number of used for data analysis because samples within quad- larvae was compared among depths and as an inter- rants were randomly located. We limited statistical action of depth and time. Time was treated as the analyses to the 19 June sampling date when the whole plot factor, depth as the subplot factor, and P. oblongus population was at its peak. The remaining sample locations within quadrants as subsamples. If F weevil species were not analyzed because of large was significant (P< 0.05), means were separated using numbers of zeroes. Fisher protected least significant difference (LSD) test with a Bonferroni correction. A square-root trans- formation was applied to fulfill the assumption of equal Results variance. Relative Abundance of Weevils. Curculionid larvae Adult Sampling. Adult occurrence in the forest un- comprised 55.2% (74.0% when the four samples con- derstory was estimated primarily by sweep net sam- taining all of the Formicidae are omitted from Hyme- pling. Sampling was conducted during 25 May through noptera) of the macro-soil insect fauna in both single 28 August 2002 and 4 June through 1 August 2003. and stratified soil samples collected in 2001 and 2002 During 2002, extensive sites were sampled approxi- (Table 2). The remainder included 25.7% Hymenop- mately monthly, and intensive sites were sampled bi- tera, 11.1%Diptera, 7.6% other Coleoptera, and 0.4% weekly. During 2003, both intensive and extensive other. In addition to insect larvae, 153 Annelida and sites were sampled biweekly during peak populations untabulated Chilopoda and Diplopoda were col- of all three species. Two randomly chosen points were lected. selected from each quadrant in each site (total = 192, The average number of weevil larvae collected per 224, and 400 for extensive and intensive sites in 2002 845-cm3 soil core sample among all sites was 0.23 2 and all sites in 2003, respectively). A sample consisting 0.05 (n = 160; range, 0 - 4) in 2001 and 0.50 + 0.06 (n = of 10 sweeps of vegetation within 1m of the forest floor 320; range, 0-10) in 2002. Of the 268 weevil larvae was collected at each point. Goodburn (1996) pro- collected, 17.2% were unidentifiable to species be- vides a complete description of understory vegetation. cause of physical condition or discrepancies from the Data collected during 2002 included coleopterans identification key (Table 3). The composition of the only; in 2003, all insects except Culicidae were in- remaining weevils was 34.3% P. sericeus, 15.7% Bary- cluded. peithes pellucidus (Boheman), 15.3%P. oblongus, 9.3% The occurrence of adults higher in the canopy was S. asperatus, 7.1% Polydrusus sp., and 1.1% Trachy- estimated by branch-beating samples. Beating samples phloeus sp. (Table 3). The Trachyphloeus sp. larvae are were collected biweekly at intensive sites. Two ran- most likely T. aristatus (Gyllenhal), because adults domly chosen trees were selected from each quadrant were collected from the site in which they were found. per site (total = 224 and 160 for 2002 and 2003, re- Polydrmsus sp. larvae could not be validated to species, spectively). One limb from each tree was sampled. A but other than P. sericeus, P. impressifions (Gyllenhal) 1-m2 ground cloth was placed under the limb before and P. americanus (Gyllenhal) are the only Polydrmsus beating. Limb height and tree species were recorded known to be prevalent throughout the Great Lakes for each sample, with limb height never exceeding region (O'Brien and Wibmer 1982, Downie and Arnett 3.5 m. In 2002, only Coleoptera were recorded from 1994, Anderson 2002). All identified curculionid larvae samples; in 2003, all insects were recorded. collected were exotic species. Emergence traps consisting of inverted 18.9-liter The diversity and abundance of all insects collected buckets were deployed at two intensive study sites from foliage are grouped by order (except for Cur- during 2002 and at one during 2003. Two glass jars culionidae, which are separated from other ) in containing tissue paper were fastened to each bucket Table 2. All insects on foliage were adult, except for 302 ENVIRONMENTALENTOMOLOGY Vol. 34, no. 2

Table 2. Total numbers of insects collected from experimental sites in northeastern Wisconsin and the upper peninsula of Michigan

Foliar samples Soil samplesa sweepingb Beating" Order1family 2001 2002 2002 2003 2002 2003

------Curculionidae 40 228 245 66 56 41 Coleoptera 12 25 13 19 5 12 Diptera 0 54 - 29 - 0 Hemiptera 0 1 - 56 - 8 Hymenoptera 0 125 - 9 - 1 Lepidoptera 0 O - 21 - 5 Mecoptera 1 0 - 1 - 0 Psocoptera 0 0 - 1 - 2

n = number of sites sampled. " Soil samples were collected September and November 2001 and May and November 2002. Data include both single auger and stratified depth samples. The number of single auger samples per date per site was 8 in 2001 and 40 in 2002. The number of stratified depth samples collected per site was 6 in 2001 and 20 in 2002. Data include sweep net samples collected mid-May through August: approximatelymonthly at n = 6 sites and biweekly at n = 4 sites during 2002, and biweekly at all 10 sites during 2003. 'Beating samples were collected approximately biweekly at n = 4 sites mid-May through August of 2002 and 2003.

Lepidoptera and some Hemiptera, and most insects in 5.5% T, aristatus. As with soil samples, all curculionids soil were immature. Additional information on the collected with sweep net, emergence trap, and beating breakdown of these orders by family can be found in techniques in stands having a predominantly sugar Pinski (2004). Curculionids comprised 94.1% of the maple understory were exotic species. Fewer species total coleopterans collected by sweep netting and were collected as adults; B. pellucidus and the un- beating in 2002. Curculionids comprised 39.5% of the determined Polydrusus sp. occurred in soil samples, total insect fauna in 2003. The remaining insects col- but not sweep net, emergence trap, or beating sam- lected by sweep net and beating (complete sampling ples. in 2003 only) were 22.1% Hemiptera, 12.9% other Sweep net sampling collected the most weevils Coleoptera, 10.7% Diptera, 9.6% larval Lepidoptera, overall (71.0%),whereas beating and emergence traps 3.7% Hymenoptera, and 1.5% other. collected 22.1 and 6.9%, respectively. Sweep net sam- Phyllobius oblongus was the predominant species pling accounted for almost all P. obhgus (82.1%) and collected from foliage, representing 66.4% of all cur- S. asperatus (100%) obtained. In contrast, sweep net culionids (Table 3). The remaining composition of sampling accounted for only 45.8% of the total weevils was 21.9% P. sericeus, 6.2% S. asperatus, and P. sericeus and beating techniques accounted for

Table 3. Numbers and species of Curculionidae collected from experimental sites in northeastern Wisconsin and the upper peninsula of Michigan

Larvae Adults 2001 2002 2002 2003 Species -- Soila Soil sweepingb Beating ~mer~ence~Sweeping Beating Emergence (n = 10) (n = 4) (n = 10) (n = 4) (n = 2) (n = 10) (n = 4) (n = 1) Barypeithes pelluciduse 6 36 0 0 0 0 0 0 Phyllobius oblonguse 22 19 190 33 0 49 19 0 Polydrusus sericezlse 5 87 33 23 5 11 22 2 Polydrrrsz~ssp. 0 19 0 0 0 0 0 0 Sciaphilus asperatus" 2 23 21 0 0 6 0 0 Trachyphloeus sp." 0 3 0 0 0 0 0 0 Trachyphloeus aristatus" 0 0 1 0 15 0 0 8 Unidentified spp. morphospecies 1 O 2 0 O 0 0 0 0 Rearing specimens 0 29 0 0 0 0 0 0 Physically deformed 5 10 0 0 0 0 0 0 Total 40 228 245 56 20 66 41 10

n = number of sites sampled. a Soil samples were collected September and November 2001 and May and November 2002, and include both single auger and stratified depth samples. The number of single soil samples collected per date per site was 8 in 2001 and 40 in 2002. The number of stratified depth samples collected per site per date was 6 in 2001 and 20 in 2002. Sweep net samples were collected mid-May through August: approximately monthly at n = 6 sites and biweekly at n = 4 sites during 2002, and biweekly at all 10 sites during 2003. "Beating samples were collected approximately biweekly at n = 4 sites mid-May through August of 2002 and 2003. Emergence trap samples were collected approximately biweekly at n = 2 and n = 1 sites mid-May through August of 2002 and 2003, respectively. " Nonindigenous species. April 2005 PINSKIET AL.:ABUNDANCE AND SEASONA~OF NONINDIGENOUS WEEVILS 303

Table 4. Temporal phenology and mean number of weevil larvae collected per 845-cm3 sample

Date Species 29 September 2001 10 November 2001 3 May 2002" 2 November 2002

Phyllobius oblongus 0.08 2 0.03 0.20 2 0.05 0.01 t 0.01 0.1 1 2 0.03 Polydrusus sericeus 0.03 -+ 0.02 0.01 2 0.01 0.20 2 0.04 0.19 2 0.05 Sciaphilus asperatus 0 0.01 2 0.01 0.03 t 0.01 0.09 2 0.05

Samples were collected from all 10 experimental sites during 2001 and from four intensive-study sites during 2002. n (single auger soil samples) = 80 for 2001 and 160 for 2002. "An additional two unidentified pupae were obtained.

46.9%. Emergence traps collected only 7.3% of the Based on combined sweep net data from 2002 (ex- total P. sericeus and no P. 0bkmgu.s or S. asperatus. cluding times when only intensive sites were sampled) However, they accounted for 95.8% of T. aristatus. and 2003,66.5% of P. oblongus adults were obtained on Seasonal Phenology of Weevils. Weevil larvae were 19 June, and all were obtained from 4 June through 16 found in the soil from 29 September through 3 May July (Fig. 1). Sixty-three percent of P. sericeus were (Table 4). The mean number of P. oblongus larvae obtained on 12 July, and all were obtained from 12 July collected per sample was 20 and 10 times, respectively, through 28 August. In contrast, only 34.6% of S. aspe- greater during the two Novembers preceding and fol- ratus were obtained within the peak sampling period, lowing the May sampling date. Conversely, the mean 12 July, and others were obtained from 4 June through number of P. sericeus larvae collected per sample did 28 August. not vary temporally at equal sampling intensities. Few Thus, all three of these species spend most of their S. asperatus larvae were collected overall, but their life cycles as larvae in the soil (Witter and Fields collection increased with successive sampling periods. 1977), but their peaks of adult activity differ. There Two curculionid pupae were collected during the May was little overlapbetween P, oblongus and P. sericeus sampling period. adults, although emergence traps indicate that P. seri-

-+- Phyllobius obkngus .. ..*-... Polydrusus sericeus - 9 - Sciaphilus asperatus

-

-

-

-

Fig. 1. Seasonal phenolog~of adult weevils collected from northern hardwood forests in northeastern Wisconsin and the upper peninsula of Michigan. Data based on sweep net samples from 10 sites collected monthly and biweekly, mid-May through August 2002 and 2003, respectively. 304 ENVIRONMENTALENTOMOLOGY Vol. 34, no. 2

N Larvae Stand Adults N

Rhinelander

Butternut Sucker 1

Taylor 1

Sylvania 2 Sylvania 6 Imp 2

Sucker 2

Tamarack 2 Taylor 2

Proportion of Weevils within a Site

I/A Phyltobius oblongus 0Polydrusus sericeus Sciaphilus asperatus

Fig. 2. Species distribution of larvae and adult weevils among northern hardwood forest sites. Proportion of species within a site is displayed, and the total number of weevils is listed after each bar grouping. Larval data are based on single auger soil samples collected from 10 sites 29 September and 10 November 2001 and from 4 intensive-study sites 3 May and 2 November 2002. Adults based on sweep net samples from 10 sites collected approximately once monthly, mid-May through August 2002.

ceus begins to emerge in late June. In contrast, there Overall, larval populations appeared to be clustered. was only a slight peak for S. asperatus adults, whose The percentage of samples that yielded zeroes on populations were more evenly distributed 4 June 29 September 2001, 10 November 2001, 3 May 2002, through 28 August, and coincided with both P. obh- and 2 November 2002 were 91.3,75.0,70.0, and 73.8%, gus and P. sericeus. respectively. Of those samples in which larvae were Distribution of Weevil Larvae Within and Between found, their numbers averaged 1.14 + 0.14,1.45 + 0.19, Sites. The average number of P. sericeus larvae (0.41 + 1.63 + 0.14, and 1.98 + 0.31 larvae1845 cm3. Even in the 0.10) within the top 10 cm of the soil surface was hybrid poplar plantation, the most homogeneous of all significantly greater than at depths of 10-20 (0.15 + habitats sampled, 42.5 and 62.5%, respectively, of the 0.05) and 20-30 cm (0.02 + 0.02) (F,,,,, = 7.07, P = 3 May and 2 November 2002 samples yielded no 0.0013).The latter two depths did not differ. The mean P. sericeus larvae, yet 7.5 and 12.5% of these samples proportion of larvae recovered at each depth was yielded 31 and 58.3% of the total (Appendix 1of Pinski similar during May, September, and November col- 2004). There is some evidence that increasing the lections (F4,im= 1.47, P = 0.2150). In six of the seven sites, one species accounted for sampling intensity may raise the overall mean of larvae 280% of total larvae, and one species accounted for collected per sample because of their aggregated dis- >44% of the total in all seven sites (Fig. 2). P. oblongus tribution. Overall, the estimates of mean larval density dominated weevil larval composition in five sites. were 44 and 95/m2 with 8 and 40 samples per site, However, the number of P. oblongus varied greatly, respectively. At the hybrid poplar plantation, for ex- with just three sites providing 75% of their total. ample, the average number of weevil larvae per soil P. sericeus and S. asperatus were predominant in one auger sample was 0.44 + 0.16 (n = 16; range, 0-2) in site each. These sites comprised 86.2% of the total P. 2001 and 1.08 ? 0.14 (n = 80; range, 0-7) in 2002. sericeus and 81% of the total S. asperatus collected. However, we cannot distinguish whether this trend is April 2005 PINSKI ET AL.:ABUNDANCE AND SEASONA~OF NONINDIGENOUS WEEVILS 305 caused by sampling intensity or between-year differ- and B. pellucidus were obtained only by larval sam- ences. pling (Table 3), although B. pellucidus have also been Distribution of Weevil Adults Between Sites. Adult collected in pitfall traps at these and similar sites weevil species composition was dominated by P. ob- (Werner and Raffa 2000). All identified weevil larvae longus in 8 of the 10 sites (Fig. 2). Their numbers are considered soil inhabitants and external rhizopha- varied greatly among sites, however, with 80.9% of the gous feeders (Kerr 1949, van Emden 1950, 1952, total collected in just five sites. P, sericeus and S. as- Vollman 1954, Brandt et al. 1996). Larval sampling peratus were predominant in one site each; 70.4% of broadened the known composition of the nonindig- the total P. sericeus were collected in just one site, and enous soil weevil complex in this region, which now 70% of the total S. asperatus were collected in just two totals a minimum of five species (Table 3). Our results sites. Thus, 8 of the 10 sites were dominated by one indicate that a relatively large number of soil samples species, accounting for >80% of the total, and all 10 are required to accurately estimate weevil abundance sites were dominated by one species, accounting for within a site. Increasing the number of samples from >60% of the total. Graphical analysis of sweep net 8 to 40 may partially address dif£iculties associated sampling data indicated that the predominant weevil with clustered larval distributions. species within each site did not vary between 2002 and Adult sampling methods varied in their applicability 2003 (Pinski 2004). However, populations were sub- among different weevil species. Sweep netting alone stantially lower in 2003 compared with 2002 (Table 3). was adequate for obtaining P. oblongus, whereas Therefore, only sweep net data from 2002 are dis- sweep netting and beating were equally effective at played to portray between-site species distributions obtaining P, sericeus (Table 3). Sweep netting ac- (Fig. 2). counted for all S. asperatus adults, but few were col- Adult weevil populations appeared to be clustered. lected overall (Table 3). In contrast, this nocturnal The abundance of P. oblongus varied significantly species comprised 50.7% of all Curculionidae col- among quadrants during the 19 June 2002 sampling at lected in pitfall traps at these and similar sites, whereas Sucker 2 (2= 29.6; df = 3; P < 0.0001), Taylor 2 P. oblongus and P. sericeus comprised only 13.5 and (2= 15.8; df = 3; P = 0.0012), and Imp 2 (2= 10.8; 0.2%, respectively (Werner and Raffa 2000). Hence df = 3; P = 0.0129) sites. The remaining sites were not comparisons of species abundance based on different analyzed statistically because of the low number of methods should be interpreted with caution. Emer- P. oblongus collected, but again, trends suggest a clus- gence traps provided the most effective method for tered distribution. The mean number of P. oblongus collecting T, aristatus, a nocturnal ground-dwelling, collected per sweep net sample across all sites on flightless, parthenogenic weevil native to Europe 19 June was 1.9 + 0.3 (n= 80; range, 0-15). Forty-five (Table 3) (Brown 1965, Suomalainen et al. 1987, Mor- percent of the samples collected no P. oblongus, ris 1997).The only prior North American records were whereas just 10%returned 51.7% (Appendix 2 of Pin- near Ottawa, Ontario (Brown 1965, O'Brien and Wib- ski 2004). Sixty-four percent of the samples that con- mer 1982) tained weevils had at least two weevils. Sweep net results may not only be affected by wee- Overall, the predominant species at sites were sim- vil behavior but also by canopy structure. The average ilar between larvae and adults (Fig. 2). This corre- number of weevils collected per sweep net sample spondence held, for example, at three of the four from even-aged stands was nearly three times greater intensive study sites. However, species compositions than from uneven-aged stands in 2002 (Pinski 2004), of weevil larvae and adults collected within sites did suggesting multilayer understories may draw adult not always coincide. For example, S. asperatus and weevils higher into the canopy. Hence, if ground P. sericeus were collected from Tamarack 2 as larvae cover is scarce, the value of branch beating increases. and from Taylor 2 as adults, but not as adults or larvae For example, most P. sericeus at the hybrid poplar from these sites, respectively. Additionally, P. sericeus plantation, where ground cover was minimal, were was collected as adults from Butternut, but not as obtained by beating (Table 3). larvae, and S. asperatus was collected as larvae from Further research should concentrate on refining Rhinelander, but not as adults. sampling protocols to employ multiple techniques (Table 3), standardization among different methods, and stratified subsampling to account for the effects of Discussion varying vegetation patterns (Table 1).Two important Exotic curculionids dominated the insect fauna in considerations are time of day when each species is both the soil and on the understory foliage, and there most active and its method of dispersal. P. oblongus and was a surprising lack of native weevils in these habitats P. sericeus are diurnally active and disperse by flight, (Table 2). Each site was dominated by a single weevil thus sweep netting and branch-beating (Table 3) species, in both larval and adult samples (Fig. 2). seem more effective than pitfall traps, which collected Overall, the two predominant species as both larvae 114 P. oblongus and only 2 P. sericeus among 23 north- and adults were P. sericeus and P. oblongus, respec- ern hardwood study sites during 1996 (Werner and tively (Table 3). Raffa 2000). Conversely, S. asperatus, B. pellucidus, Despite the added labor, soil sampling is a valuable and T. aristatus are nocturnally active, flightless, leaf- component of assessing overall weevil species com- litter inhabiting weevils (Brown 1965, Galford 1987, position. For example, the unidentified Polydrusus sp. Suomalainen et al. 1987, Morris 1997), so they were 306 ENVIRONMENTALENTOMOLOGY Vol. 34, no. 2 less commonly collected in our sweep net samples Brown, V. K., and A. C. Gange. 1992. Secondary plant suc- (Table 3) than previous pitfall trap samples, which cession: how is it modi£ied by insect herbivory?Vegetatio. yielded 428 S. asperatus and 225 B. pellucidus (Werner 101: 3-13. and Raffa 2000). Brown, W. J. 1940. Notes on the American distribution of The absence of native weevil species in both un- some species of Coleoptera common to the European and derstory vegetation and forest soils throughout this North American continents. Can. Entomol. 72: 65-78. Brown, W. J. 1965. Trachyphloeus Gennar (Coleoptera: region is surprising, given the frequency, distribution, Curculionidae) in North America. Can. Entomol. 97: 189- and intensity of our sampling efforts. Moreover, ob- 192. servations by Witter and Fields (1977) report similar Carruth, L. A. 1936. Further observations on the oblong leaf findings. The most likely explanation is competitive weevil in western New York. J. Econ. Entomol. 29: 806- displacement by these five nonindigenous weevils, 807. particularly P. oblongus and P. sericeus. However, it is Casteels, H., and R. De Clercq. 1988. Observations on wee- difficult to resolve such discrepancies or document vil pests in nurseries. Meded. Fac. Landbouwwet. Rijks- competitive displacement without baseline biodiver- univ. Race. Int. Symp. Crop Prot. 53: 1169-1174. sity inventories (Kremen et al. 1993, Kremen 1994, Czerniakowski, Z. W. 1998. The composition of weevils (Coleoptera: Curculionidae) species occurred on willow Sharkey 2001). plantations in southeastern Poland. Prog. Plant Prot. 38: 160-164. Downie, N. M., and R. H. Arnett. 1994. The beetles of Acknowledgments Northeastern North America, vol. 2. Sandhill Crane Press. Gainesville, FL. Field and laboratory assistance by J. Ludden, A. Boyd, Felt, E. P. 1928. Observations and notes on injurious and E. Lewandowski, K. Kieler, and J. Frie, Department of En- other insects of New York State. N. Y. State Mus. Bull. 274: tomology, University of Wisconsin (UW)-Madison and spec- 145-76. imen identificationby S. Krauth, Department of Entomology, Fields, R. D. 1974. Weevil feeding on sugar maple repro- UW-Madison, and C. O'Brien, Center for Biological Control, duction in the Upper Peninsula of Michigan. MS thesis, A & M University, was greatly appreciated. T. Steele, Kemp University of Michigan, Ann Arbor, MI. Natural Resources Station, assisted with collection of pre- Frost, C. A. 1946. Polydrusus sericacs Schall. Psyche. 53: 51. liminary data and operational support. E. Latty, Biology De- Galford, J. R. 1987. Feeding habits of the weevil Barypeithes partment, Hollins University, provided valuable site data. We pellucidus (Coleoptera: Curculionidae). Entomol. News. thank B. Aukema, Biometry Program, UW-Madison, for help 98: 163-164. with statistical analyses. We thank USFS personnel from the Gharadjedaghi, B. 1997. Phytophagous fauna on Sylvania Wilderness Area, Nicolet National Forest, and Ot- alders (Alnus spp.) in banks of brooks in upper Frankonia tawa National Forest for providing research sites. The critical Part 1: Assessment by the beating method. Forstwiss. reviews of D. Young, D. Mahr, Department of Entomology, Centralbl. 116: 158-177. UW-Madison, E. Kruger, Department of Forest Ecology and Goodburn, J. M. 1996. Comparison of forest habitat struc- Management, UW-Madison, and two anonymous reviewers ture and composition in old-growth and managed north- are greatly appreciated. The USDA Forest Service, RJVA ern hardwoods in Wisconsin and Michigan. MS thesis, 00-JV-11231300 - 090, and the UW-Madison, College of Ag- University of Wisconsin, Madison, WI. ricultural and Life Sciences, funded these studies. Goodburn, J. M., and C. G. Lorimer. 1998. Cavity trees and coarse woody debris in old-growth and managed north- em hardwood forests in Wisconsin and Michigan. Can. J. For. Res. 28: 427- 438. References Cited Goodburn, J. M., and C. G. Lorimer. 1999. Population struc- Anderson, R. S. 2002. Curculionidae, pp. 722-815. In R. H. ture in old-growth and managed northern hardwoods: an Arnett, M. C.Thomas, P. E. Skelley,and J. H. Frank (eds.), examination of the balanced diameter distribution con- American beetles, vol. 2. : Scarabaeoidea cept. For. Ecol. Manag. 118: 11-29. through Curculionoidea. CRC, Boca Raton, F'L. Haack, R. L., and T. M. Poland. 2001. Evolving management Axelsson, B., S. Larsson, U. Lohm, and T. Person. 1973. strategies for a recently discovered exotic forest pest: the Production of larvae of Phyllobius piri L. (Coleoptera: pine shoot beetle, Tomicm piniperda (Coleoptera) . Biol. Curculionidae) in an abandoned field. Entomol. Scand. Invasions. 3: 307-322. 4: 21-29. Hain, F. P., R. G. Hollingsworth, F. H. Arthur, F. Sanchez, Blatchley, W. S., and C. W. Leng. 1916. Rhynchophora or and R K. Ross. 1991. Adelgid host interactions with spe- weevils of North Eastern America. Nature Publishing, cial reference to the balsam woolly adelgid in North Indianapolis, IN. America. Gen. Tech. Rep. NE 153: 271-287. Bockheim, J. G. 1997. Soils in a hemlock-hardwood ecosys- Helsen, H., and L. Blommers. 1988. Toxicity of Evisect S to tem mosaic in the southern Lake Superior uplands. Can. apple insects. Meded. Fac. Landbouwwet. Rijksuniv. J. For. Res. 27: 1147-1153. Race. Int. Symp. Crop Prot. 53: 979-986. Borror, D. J., C. A. Triplehorn, and N. F. Johnson. 1989. An Henshaw, S. 1888. Two European weevils new to the United introduction to the study of insects, 6th ed. Saunders States. Psyche. 5: 137. College Publishers, Philadelphia, PA. Hoffman, A. 1950. Faune de France. vol. 52, Coleopteres Brandt, J. P., S. M. Smith, and M. Hubbes. 1996. Distribution curculionides, Premiere partie. Federation francaise and sampling of root weevil larvae in young ornamental des societes de sciences naturelles, Office Central de conifer plantations. Can. Entomol. 128: 1125-1133. Faunistique, Paris, France. Brown, V. K., and A. C. Gange. 1991. Effects of root her- Hunter, M. D. 2001. Out of sight, out of mind: the impacts bivory on vegetation dynamics. Spec. Publ. Br. Ecol. Soc. of root-feeding insects in natural and managed systems. 10: 453-470. Agric. For. Entomol. 3: 3-9. April 2005 PINSKIET AL.: ABUNDANCEAND SEASONALITYOF NONINDIGENOUS WEEVILS 307

Kerr, T. W., Jr. 1949. The arborvitae weevil, Phyllobius in- Parrott, P. J., and H. Glasgow. 1916. The leaf weevil (Poly- trusus Kono. Bull. 305. Rhode Island Agricultural Exper- drusus impressifions Gyll.) . Technical Bulletin. New York imental Station, Kingston, RI. Agricultural Experiment Station. Kremen, C. 1994. Biological inventory using target taxa: a Pimentel, D., L. Lach, R. Zuniga, and D. Morrison. 2000. case study of the butterflies of Madagascar. Ecol. Appl. 4: Environmental and economic costs of nonindigenous 407- 422. species in the United States. BioScience. 50: 53-65. Kremen, C., R. K. Colwell, T. L. Erwin, D. D. Murphy, Pinski, R. A. 2004. Biology and host range of an invasive R. F. Noss, and M. A. Sanjayan. 1993. Terrestrial arthro- root-feeding weevil complex affecting northern hard- pod assemblages: their use in conservation planning. Con- wood forests. MS thesis, University of Wisconsin, Madi- sew. Biol. 7: 796-808. son, WI. Lawrence, J. F. 1991. Key to the families and many subfam- Sala, 0. E., F. S. Chapin 111, J. J. Armesto, E. Berlow, J. ilies of Coleoptera larvae (Worldwide), pp. 184-296. In Bloomfield, R. Dirzo, E. Huber-Sanwald, L. F. Huen- F. W. Stehr (ed.), Immature insects, vol. 2. Kendall-Hunt neke, R. B. Jackson, A. Kinzig, R. Leemans, D. M. Lodge, Publishing, Dubuque, IA. H. A. Mooney, M. Oesterheld, N. L. Poff, M. T. Sykes, B. Levesque, C., and G. Y. Levesque. 1994. Abundance and H. Walker, M. Walker, and D. H. Wall. 200. Global seasonal activity of weevils (Coleoptera: Curculionidae) biodiversity scenarios for the year 2100. Science. 287: in a raspbeny plantation and adjacent sites in southern 1770 -1774. Quebec (Canada). Gt. Lakes Entomol. 27: 23-37. SAS Institute. 1999. SAS OnlineDoc, ver. 8. SAS Institute, Liebhold, A. M., W. L. MacDonald, D. Bergdahl, and Cary, NC. V. C. Mastro. 1995. Invasion by exotic forest pests: a Sharkey, M. J. 2001. The all taxa biological inventory of the threat to forest ecosystems. For. Sci. Monogr. 30: 1-49. Great Smoky Mountains national park. Fla. Entomol. 84: Lodge, D. M., and K. Shrader-Frechette. 2003. Nonindig- 556 -564. enous species: ecological explanation, environmental eth- Simmons, G. A., and F. B. Knight. 1973. Deformity of sugar ics, and public policy. Conserv. Biol. 17: 31-37. maple caused by bud feeding insects. Can. Entomol. Mack, R. N., D. Simberloff, W. M. Lonsdale, H. Evans, 105: 1559 -1566. M. Clout, and F. Bazzaz. 2000. Biotic invasions: causes, Sleeper, E. L. 1957. Notes on North American species of epidemiology, global consequences, and control. Ecol. Polydrusus Germar (Coleoptera: Curculionidae, Brachy- Appl. 10: 689-710. derinae). Ohio J. Sci. 57: 129-134. Massee, A. M. 1932. Notes on insect pests and mites in 1932. Suomalainen, E., A. Saura, and J. Lokki. 1987. Cytology and Ann. Rep. East Malling Res. Stn. 20: 109-116. evolution in parthenogenesis. CRC, Boca Raton, FL. Massee, A. M. 1954. Pests of fruits and hops. Crosby Lock- Syvertsen,J. P., and C. W. McCoy. 1985. Leaf feeding injury wood and Son, London, UK. to citrus by root weevil adults: leaf area, photosynthesis Mattson, W. J., P. Niemela, I. Millers, and Y. Inguanzo. 1994. and water use efficiency. Fla. Entomol. 68: 386-393. Immigrant phytophagous insects on woody plants in the van Emden, F. I. 1950. Eggs, egg-laying habits and larvae of United States and Canada: an annotated list. U.S. Depart- short-nosed weevils. Proc. 8th Int. Congr. Entomol. ment of Agriculture, Forest Service, North Central Forest Stockholm, Sweden. Experiment Station, St. Paul, MN. van Emden, F. I. 1952. On the Rhynchophora larvae: Ad- Melure, M. S., and C.A.S.-J. Cheah. 1999. Reshaping the elognatha and Alophinae (Insecta: Coleoptera) . Proc. ecology of invading populations of hemlock woolly ad- Zool. Soc. Lond. 122: 651-795. elgid, Adelges tsugae (Homoptera: Adelgidae) ,in eastern Vitousek, P. M., C. M. D7Antonio,L. L. Loope, and R. West- North America. Biol. Invasions. 1: 247-254. brooks. 1996. Biological invasions as global environmen- Ministry of Agriculture, Fisheries, and Food. 1963. Wing- tal change. Am. Sci. 84: 468-473. less weevils. Advisory Leaflet 57. Ministry of Agriculture, Vollman, M. 1954. Der schmalbauchriissler Phyllobius ob- Fisheries, and Food, Edinburgh, Scotland, UK. Eongus L. (Coleoptera: Curculionidae) Ein Beitrag zur Morris, M. G. 1978. Polydrustls sericeus (Schaller) (Co- Biologie und Bekampfung. Z. Angew. Entomol. 36: 117- leoptera: Curculionidae) in south Wiltshire. Entomol. 155. Rec. J. Var. 90: 22-23. Werner, S. M., and K. F. Raffa. 2000. Effects of forest man- Morris, M. G. 1997. Broad-nosed weevils (Coleoptera: agement practices on the diversity of ground-occurring Curculionidae) () . Handbooks for the identi- beetles in mixed northern hardwood forests of the fication of British insects, vol. 5, part 17a. Royal Ento- Great Lakes Region. For. Ecol. Manag. 139: 135-155. mology Society, London, UK. Wilcove, D. S., D. Rothstein, J. Dubow, A. Phillips, and Nowak, D. J., J. E. Pasek, R. A. Sequeira, D. E. Crane, and E. Losos. 1998. Quantifying threats to imperiled species V. C. Mastro. 2001. Potential effect of Anopbphora gla- in the United States. BioScience. 48: 607-615. brtpennis (Coleoptera: Cerambycidae) on urban trees in Witter, J. A., and R. D. Fields. 1977. Phyllobius oblongus and the United States. J. Econ. Entomol. 94: 116-122. Sciaphilus asperatus associated with sugar maple repro- 07Brien,C. W., and G. J. Wibmer. 1982. Annotated check- duction in northern Michigan. Environ. Entomol. 6: 150- list of the weevils (Curculionidae sensu lato) of North 154. America, Central America, and the West Indies (Co- leoptera: Curculionidae). Memoirs of the American En- Received for publication 28 June 2004; accepted 28 Decem- tomological Institute, Ann Arbor, MI. ber 2004.